Contextual Power Rating

[virgil]

How transparent are the numbers and the sources of their particular CPR going to be with the players?

I've postulated on these boards before on something akin to this, but nothing so pervasive. Something more akin to giving a monster more than one CR, based on its 'context' (troll in a small room, troll in an open field). Maybe even an ECL modifier for the player and whether its specialization applies.

[Roy]

Ugh.

You manage to raise not one single point I haven't already addressed and/or debunked.

I am especially offended by the "counteracts intelligent play thing". I addressed that in fucking detail and you are talking out your fucking ass by harping on over something I've debunked without for a second addressing my material debunking it.

If you are going to be like that then fuck off out of my thread.

So where did you address shit like PCs beating themselves to make the mobs easier, and avoiding playing smart by using actual tactics and such so the mobs don't get harder?

[Orion]

PhoneLobster actually did address that.

He said CPR is a diagnostic tool and that he hasn't yet made a concrete proposal for its actual application.

So if you *want* the PCs to be able to get an advantage by choosing to exploit their conditional abilities, you can simply *let* them go against lower-CPR challenges. You don't have to make it harder, but at least CPR tells you how much easier it is.

[Anguirus]

So where did you address shit like PCs beating themselves to make the mobs easier, and avoiding playing smart by using actual tactics and such so the mobs don't get harder?

For PC's beating themselves to make mobs easier, ideally CPR is balanced well enough that these sorts of things aren't exploitative because, while the challenges become easier on an absolute level, they remain exactly as difficult for the participants to deal with as they would be had they not been injured, you're just less able to deal with unfavorable extremes of probability because you have fewer degrees of nuance in tracking your health. That is to say, while you have fewer health points and your opponents attack output is designed to correlate to the number of health points that you have, however, shifts in damage outputs in a scenario where fewer health points are representing active and inactive are going to be disproportionately damning. There doesn't appear to be any reason to go into a fight with less health than you absolutely could have in such a system but there do appear to be advantages to having more health. And just to preempt a response of "but if you have less health then they should have less health" I would say no, not necessarily. I don't know what Phone Lobster has in mind but it could just as easily be that an opponent's health is dependent on how much damage output you do and their attacks are dependent on how much health you have. In such a situation, I think, you may end up with it being beneficial to not do a lot of damage and I'm curious how Phone Lobster deals with that.

[Roy]

So that's 4.Fail on crack, where you're always facing the same success rate against auto scaling mobs.

[RandomCasualty2]

For PC's beating themselves to make mobs easier, ideally CPR is balanced well enough that these sorts of things aren't exploitative because, while the challenges become easier on an absolute level, they remain exactly as difficult for the participants to deal with as they would be had they not been injured, you're just less able to deal with unfavorable extremes of probability because you have fewer degrees of nuance in tracking your health.

Yeah, assuming CPR errs on the side of giving full strength PCs more of a benefit, then it won't be a problem. Ridiculous tactics will only get used if the PCs actually gain something from using them, and in the case of having weaker HPs, this only increases the rocket launcher tag aspect of combat, which is inherently bad for PCs even if all other factors are even.

The problem here is that the system really doesn't reward good tactical play. Pretty much regardless of what you do, you can't improve or hurt your chances. To some degree that's not a bad thing, because it allows different play styles to work, but you have to be careful to actually punish really stupid actions and give PCs some reward for doing creative things. But I am actually kind of okay saying that you can ramp up the opposition if the PCs do something really cheesy like scry & die prebuffing.

[PhoneLobster]

I don't know what Phone Lobster has in mind but it could just as easily be that an opponent's health is dependent on how much damage output you do and their attacks are dependent on how much health you have. In such a situation, I think, you may end up with it being beneficial to not do a lot of damage and I'm curious how Phone Lobster deals with that.

It's nice to see that most people are beginning to grasp the whole measuring stick that tracks variation in power rather than iron clad rule of blandness thing.

As for damage output and health.

OK so CPR in my example system rather specifically measures attack and defense of opponents separately and makes comparisons of each sides attack values against their weakest opposing defense values.

And if your GM decided to use those values to better balance to the parties capabilities he has the information to balance opponents attack against the parties defense and opponents defense against the parties attack. This is pretty important because unless we provide additional character build rules that restrain character advancement so that attack and defense advance together you kind of need this information to use CPR as a substitute balance mechanism for it.

Being a character who leans towards defense is no more beneficial than any other strategy, the CPR system I've outlined having effectively taken into account attack/defense variation.

As to being injured or running out of ammo. CPR accounting for it does NOT enforce changes to opposition, people are getting that, if you stab yourself the GM can measure the change to CPR and laugh in your face rather than account for it.

Or he CAN account for pre-existing injuries and depleted resources so if he sends you on the "urgent adventure of no resting at all" you aren't specifically punished for it compared to the "5-minute work day adventure", which in an average encounter game balanced on the assumption of one, the other, or some proportion of both you WILL be.

[cthulhu]

I like the idea - accounting might be a bit complex, but it would work well in a computer supported enviroment

Can you simplify the assessment of party statE?

[PhoneLobster]

I like the idea - accounting might be a bit complex, but it would work well in a computer supported enviroment

Can you simplify the assessment of party statE?

The accounting method of my current implementation is a bit complex, but comes in at approximately one round of combat accounting.

You could readily simplify the system further in all sorts of ways and end up with single value stuff a lot like CR, but the more you simplify it the less information it provides and the less accurate the information.

You could simplify the assessment of party state, but do you mean simplify or explain?

[cthulhu]

Nah, just simplify.

[Draco_Argentum]

So that's 4.Fail on crack, where you're always facing the same success rate against auto scaling mobs.

Gah, hes repeatedly said its a measuring stick.

Being a character who leans towards defense is no more beneficial than any other strategy, the CPR system I've outlined having effectively taken into account attack/defense variation.

I disagree, you've stated its a perfect knowledge system. That means I know which of the enemies to target, hint, its not the uber defense guy. Having one team-member with massive defense doesn't help a lot. I don't know your system well enough to say for sure but its a good bet that teams with large variations in attack and defense between members will be much less effective for a given CPR than teams where everyone has a similar attack and defense.

I think you'll have to go to plan B too. The stun attack example isn't really a binary where you can be at zero CPR or full. Its entirely dependant on the frequency stuns are laid down.

Another similar issue: say I have and attack that does 5 damage and a second attack that stuns. Say a second character has an attack that does 5 damage and a second attack that does 30 damage to stunned targets. The best strategy is pretty clear, I'm assuming you wouldn't charge full CPR for the 5 damage attacks that won't get used. I think conditional CPR accounting is where slowness would arise. Unless you wrote a spreadsheet or simple java app to crunch the numbers.

[PhoneLobster]

I disagree, you've stated its a perfect knowledge system. That means I know which of the enemies to target, hint, its not the uber defense guy.

And my CPR system already makes as one of its base assumptions that attackers will always target the target that is weakest to their attacks.

I assume the specific question he was asking was that since characters with higher defense ratings aren't especially accounted for by my current CPR metric (because of that assumption) that it was like getting free power or something. But really that only works if you bring along weaker defense characters to cop it for you and lower your CPR ratings on defense.

CPR as implemented in my example provides the tools to expand your analysis of a encounter match up. So if players try and bring peasants to every encounter so their defense gets underated first of all CPR lets you know exactly what is happening and you can use the tools presented by CPR to rate the players rather than the peasants by changing a base assumption of CPR and instead assuming the peasants won't be primary targets.

Having one team-member with massive defense doesn't help a lot.

Which is why I didn't make my CPR system estimate assuming defenders favor or anything. I don't regard the ability of an ally to survive as especially important to a weaker character who wants to survive themselves. And I'm more concerned with having the system accurately predict the first casualties so balancing points can be more easily tuned on about a 1 or less PC going down sort of level.

I don't know your system well enough to say for sure but its a good bet that teams with large variations in attack and defense between members will be much less effective for a given CPR than teams where everyone has a similar attack and defense.

A team with high variation in defense probably does better than rated (as a team) on the survival front, but not necessarily on their ability to deal damage depending on whether the biggest damage contributors are the weakest defenders or not.

But a team with high variation in attack isn't especially less predictable, because in theory CPR as implemented here has measured each attackers CPR individually.

I think you'll have to go to plan B too.

Well I don't have to it's an accuracy vs complexity issue. You can make some assumptions and just cop the accuracy loss if you want.

Remember the most accurate form of CPR involves running through the numbers of all potential combat outcomes in a process that would take significantly longer to resolve than the combat itself. Everything that makes CPR more accessible and usable in play makes it less accurate, it's just a question of whether and where an appropriate balance exists.

Another similar issue: say I have and attack that does 5 damage and a second attack that stuns. Say a second character has an attack that does 5 damage and a second attack that does 30 damage to stunned targets. The best strategy is pretty clear, I'm assuming you wouldn't charge full CPR for the 5 damage attacks that won't get used.

Well the "best attacks" part of the CPR accounting, the bit where you decide which selectable abilities (like distinct types of attack actions) are counted isn't written into the CPR condition of the ability.

Effectively "+4 attack Vs Giants" on particular attack has the CPR value of "+4 attack Vs Giants" (yeah, pretty direct really). IF you are using an attack option with that bonus attached, that is it's conditional CPR value.

But the act of deciding that CPR is using that attack option is instead a step in the CPR process determined by the attackers favor/all out attack assumptions I've implemented.

And it's the same operation players use in combat resolution time in a normal game to determine which attacks they are using anyway. Now you can say "Therefore it must be pretty easy/fast", which is good. Or you can say "hold up, players only really do that once or twice a combat and re-use the information for the rest" in which case I point out since the CPR calculation I'm using just happens to in it's process perform that operation and give you your actual practical attack bonus you haven't actually increased the number of initial times per encounter that this operation is executed.

I think conditional CPR accounting is where slowness would arise. Unless you wrote a spreadsheet or simple java app to crunch the numbers.

Actually I think the conditional bit is going to be the fastest. It is, especially in my example system, the same process players use regularly and quickly in combat as it is. If you consider the number of conditional bonuses that can apply or not, and even change on a turn by turn basis in a 3.x edition D&D game the ability of players to whip through a list of conditions and tally up a total bonus is pretty sharp.

The bit I'm concerned about for complexity, and which is why I didn't go Plan B all the way is taking average damage per turn off a table (the table is the easy bit) then multiplying it by various values. Because THAT is an operation that isn't regularly performed as part of combat, and is an operation that players may well find harder than simply performing a sum of applicable bonuses.

Implement multiplications or fractions and stuff too extensively and then you WILL have a CPR system that requires computer assistance to calculate in reasonable time.

[Username17]

Ugh.

If you are going to be like that then fuck off out of my thread.

That's fair. I don't like your system, and I will fuck off out of the thread. But first, I am going to go on a tiresome rant to attempt to explain why your system doesn't do what I want it to do. Hopefully we'll understand each other better and come to loggerheads less often in the future.

---

First of all, I am very much in favor of Shadowrun style challenges. Where the opposition is persistent and capable on many levels and it is up to the players to make a strategy that utilizes what talents they have to effect victory. A system in which the method of player attack is inflexible or obvious is of no use to me whatsoever. Perfect information systems therefore are way too railroady for my tastes. I'm never going to be satisfied with something like that.

But beyond that, I genuinely value abstract average challenge ratings. They serve a genuine purpose. I want people to come up with ways to take out powerful enemies in clever ways that are less instantaneously dangerous. And I don't think the prestige of the accomplishment is any less just because you did it the easy way rather than the hard way. In short, to judge the value of the player's accomplishments, CR is a more valuable number than CPR. I want to judge a character's defeat of a minotaur. I don't hold it against them that they figured out how to pepper the minotaur with arrows from a high cliff or slit its throat in its sleep. In fact, good for them.

The only thing I would want from a CPR system is a way to determine whether a prospective encounter was likely to be a TPK. That would be good information to have. But that's not really the direction you're going with this. CPR as you're currently formatting it is pretty much a way to estimate whether the first character on your side will drop before the first character on the other side drops. That's interesting information I guess, but it's frankly not that important to me since often times teams with medics can have several characters drop early in a combat without having that ever seriously endanger their chances of victory. I only want to know whether the entire team is likely to go down before they achieve victory or whether individual characters are likely to permanently die (rather than being incapacitated). And for all the math being thrown around, I am unconvinced that your proposal will give me either of those pieces of information any better or faster than just wargaming the actual battles ahead of time.

So no, I'm not interested in your system. It is not something which I would use, because while I would not say that what it does is objectively bad, I will say categorically that it does not do what I want it to on any level. And now, I'll go fucking off now. But I hope that I've been able to elucidate our differences in a clear enough way that you and I don't have to shout at one another.

-Username17

[PhoneLobster]

First of all, I am very much in favor of Shadowrun style challenges. Where the opposition is persistent and capable on many levels and it is up to the players to make a strategy that utilizes what talents they have to effect victory.

And this is why I am so pissed with your argument style lately. Total and utter inconsistency bordering on intellectual dishonesty.

Just the last time I was posting around here you went through an entire fucking thread arguing directly against that with your Farmer vs Sailor scenario and demands for 2-dimensional "non specialists with no general abilities".

Now I deliver something that, whatever its potential pit falls, is the only thing remotely capable of answering your Farmer vs Sailor scenario... well now you actually want something different...

Are you now going to decide you were wrong on the other thread and that a Unified character system is what is required to deliver rich and rewardingly complex NPCs that don't rob your games of value by taking up 60% of the rules interactions with a kindergarten version of the PC rules?

Perfect information systems therefore are way too railroady for my tastes. I'm never going to be satisfied with something like that.

What the hell? Should I believe that statement? I'm pretty sure you have argued vociferously against Gygaxian secret rules and Cthuloid puzzle monsters against RCs preferences.

Why should I believe this claim is anything more than a momentary whimsical adoption being used to support whatever your latest whim of game design might be? Just waiting to be trashed as bad for gaming made into life incarnate the second you have some OTHER conflicting design decision you want to justify.

I want people to come up with ways to take out powerful enemies in clever ways that are less instantaneously dangerous.

Measuring stick, bitch, stop complaining how knowing about your decisions makes them not exist.

CPR as you're currently formatting it is pretty much a way to estimate whether the first character on your side will drop before the first character on the other side drops.

Pretty close. But that's not a bad point from which to extrapolate a likelihood of victory for an entire encounter, and far more accurate and tailored than average encounter for level type predictions like CR, they don't even really have even close to as clearly a defined or likely to succeed means of making their estimates.

Also considering that most NPC groups are either homogeneous or majority homogeneous with the odd boss or something the estimates will both tend to be more accurate on one side of the equation and where diversity errors favor anyone falling out in favor of the PCs.

But I hope that I've been able to elucidate our differences in a clear enough way that you and I don't have to shout at one another.

I appreciate the effort, but when you go doing things like declaring that Antigravity is just plain OK while, Antigravity is unrepresentable and bad for game play...

... expect me to point and laugh and call you names at great length.

And to keep reminding you about it for fucking forever. Or at least until you get it bloody right.

[Anguirus]

My point earlier was that having more degrees of nuance, as it pertains to hit points, makes you (whoever you are) less susceptible to the extremes of probability. This, in turn, is very good for survivability because you aren't ever going to be one shoted and resources are spent to their fullest. That is, if given the option of playing a character with 10 health that takes damage in increments of d10s or a character with 100 health that takes damage in increments of d100s (or even better 10d10's) it is clearly better to go with the 100, even though the average % of health lost per round is the same simply because the odds of being killed in one blow are much less and your ability to use a heal spell to its fullest potential is greater. If your opponents health automatically scales to how much damage you do per round then you are in a better position doing less damage per round because your opponent will have less health and thus be subject to the limitations presented above. Does this make sense? Am I misunderstanding how CPR works?

[PhoneLobster]

You are effectively making the measuring stick mistake again.

You aren't automatically scaling health to damage. You are doing it manually.

CPR doesn't actually give you anything. It just lets the GM determine if he wants to give you something.

[Anguirus]

Right, but if you as a DM send your PC's a threat that they cannot deal with because you don't have a diagnostic tool to determine if it is an appropriate threat that's one thing. If you send them the same threat knowing full well it will kill them that's another. I know that this example assumes hostility, or at least a lack of cooperation from the players and that group is doomed from the start but CPR allows shitty players to hold your game hostage.

[PhoneLobster]

You aren't arguing from a sensible position. Either you complain because CPR prevents TPKs or you complain because it fails to do so.

Trying to argue both at once is just stupid.

[RandomCasualty2]

I know that this example assumes hostility, or at least a lack of cooperation from the players and that group is doomed from the start but CPR allows shitty players to hold your game hostage.

Well, there's no way to account for shitty players. CPR might help people playing shitty classes, but as far as shitty players it won't. Because you're still going to get points for having sneak attack on your list, even if you're too dumb to find ways to use it effectively.

There's no way to account for people who have the necessary tools available and just don't use them effectively, aside from just using weaker encounters.

The actual premise of the CPR system isn't bad. The big question however is if it can be done quickly and that I just don't think is possible.

Another issue is how CPR is going to handle things that are partially effective. For instance if I take fireball and half of the monsters have fire resistance, how much credit do I get for it CPR wise? This is especially a factor with area spells because their main power comes in their ability to hit multiple foes. So if most of the foes are immune or highly resistant, they become much weaker tactical choices.

[Draco_Argentum]

A team with high variation in defense probably does better than rated (as a team) on the survival front, but not necessarily on their ability to deal damage depending on whether the biggest damage contributors are the weakest defenders or not.

Frank sort of mentioned what I was getting at with this line " I only want to know whether the entire team is likely to go down before they achieve victory or whether individual characters are likely to permanently die (rather than being incapacitated)." but didn't really go into detail.

The problem is that if a team has an obvious glass cannon compared to the rest of the team that dude is going down first. When that happens the fight is suddenly very different from the fight the CPR was supposed to predict. I'm not convinced knowing the first to drop is a meaningful predictor of party success.

Glass cannons are probably a bad idea in general IMO. Being the first to get booted to Smash Bros. every combat isn't cool.

[PhoneLobster]

Another issue is how CPR is going to handle things that are partially effective. For instance if I take fireball and half of the monsters have fire resistance, how much credit do I get for it CPR wise? This is especially a factor with area spells because their main power comes in their ability to hit multiple foes. So if most of the foes are immune or highly resistant, they become much weaker tactical choices.

Currently the idea is the bit where it's a fire attack checks against the weakest targets for it's CPR value.

The bit where it's a multi target attack checks if there are multiple targets in order to add it's CPR value.

If those targets are some proportion of targets that aren't the targets weakest to the attack then somewhere along the line there is effectively information that CPR is just not accounting for.

But CPR cannot go accounting for everything accurately. It has to take short cuts, you have to maintain simplicity discipline or it will rapidly balloon into genuine complexity issues.

The weakest target to fire attack and the multi target conditional checks are a good enough approximation, to actually account for that rather elaborate (but not entirely uncommon) issue would involve far more elaborate accounting.

I'm prepared to just call this an area that my implementation of CPR fails to fully account for, it's just not worth the expense.

[Draco_Argentum]

The actual premise of the CPR system isn't bad. The big question however is if it can be done quickly and that I just don't think is possible.

I think it can be, but I'd mostly be interested in a system like this as a replacement for CR. So its something a DM would use between games in general. If the PCs do throw the DM a curveball they need to be a little understanding when it takes the DM longer to respond than normal.

[PhoneLobster]

Glass cannons are probably a bad idea in general IMO. Being the first to get booted to Smash Bros. every combat isn't cool.

Yeah, glass cannons kind of suck. I'm not especially keen on having a glass cannon "striker" role on the PC front. But defining that as in or out that is a role for other aspects of the system, CPR can (to a limited degree) accommodate glass cannons, though certainly if that was part of you master plan you might want to design your CPR implementation based on different assumptions and metrics.

But for me having a player twiddling their thumbs for an excessive period because they got knocked out of combat first is a problem in any circumstance. So I want to balance my combats around knowing if it is going to happen and then if it does how close to finishing it the team is when it does (and also to which character it is going to happen).

My current setup up favors delivering information of a nature and in a format that would help with that.

If you want a balance point where combats are frequently lost outright by the more diverse group, or where the PCs regularly see the majority of the party go down and wander off to argue about star wars in the next room then you will want to change your CPR assumptions and metrics.

[Fuchs]

I hate long combats anyway. I don't want to grind an enemy down.

(in Gamer terms: I prefer FPS style to MMOG style)

[RandomCasualty2]

I think it can be, but I'd mostly be interested in a system like this as a replacement for CR. So its something a DM would use between games in general. If the PCs do throw the DM a curveball they need to be a little understanding when it takes the DM longer to respond than normal.

The problem is that this cannot be used between games, because it's designed to account for stuff like wounds, expended spell slots and choices of prepared spells. And when you're basing something off a party's current condition, it has to be done constantly as the game is running.

Another problem I just thought of is how CPR is going to handle scenarios where monsters have radically different DCs to defeat. For instance, a pair of giant guards may be tough combatants, but their perception is low, so sneaking by them while invisible is a suitable way to defeat the encounter. Do you just assign huge bonuses to a PC who took mass invisibility or what?